huddy

8/9/2019 Huddy

1/30

From Social to Political Identity: A CriticalExamination of Social Identity Theory

Leonie HuddyState University of New York at Stony Brook

Interest in the concept of identity has grown exponentially within both the humanities and social sciences, but the discussion of identity has had less impact than might be expected onthe quantitative study of political behavior in general and on political psychology morespecifically. One of the approaches that holds the most promise for political psychologistsis social identity theory, as reflected in the thinking of Henri Tajfel, John Turner, and colleagues. Although the theory addresses the kinds of problems of interest to political

psychologists, it has had limited impact on political psychology because of social identity

theorists disinclination to examine the sources of social identity in a real world complicated by history and culture. In this review, four key issues are examined that hinder the successfulapplication of social identity theory to political phenomena. These key issues are theexistence of identity choice, the subjective meaning of identities, gradations in identitystrength, and the considerable stability of many social and political identities.

KEY WORDS: social identity, identity politics, political identification, intergroup relations.

Interest in the concept of identity has grown exponentially during the lastdecade or so within both the humanities and social sciences. Postmodern theorists

in the humanities have challenged traditional conceptions of identity by arguingthat the fixed subject of liberal humanistic thinking is an anachronism that shouldbe replaced by a more flexible individual whose identity is fluid, contingent, andsocially constructed (Butler, 1990; Novotny, 1998; Villancourt Rosenau, 1992;Young, 1997). Social scientists have also intensified their longstanding interest inthe concept of identity in recent years (Jenkins, 1996). Sociologists have ponderedand explored the tension between individual identity and the constraints of socialstructure (Giddens, 1991; Jenkins, 1996; Stryker, 1980). Anthropologists haveexamined the cultural expression of identity, its meanings, and how it is maintained

at group boundaries (Barth, 1969; Cohen, 1986). Social psychologists have focused

Political Psychology, Vol. 22, No. 1, 2001

127

0162-895X 2001 International Society of Political PsychologyPublished by Blackwell Publishers, 350 Main Street, Malden, MA 02148, USA, and 108 Cowley Road, Oxford, OX4 1JF, UK.

8/9/2019 Huddy

2/30

on the multifaceted and situationally contingent nature of individual identity(Gergen, 1971; Hogg, Terry, & White, 1995; Markus, 1977). They have alsoidentified social identity as a powerful ingredient in the development of ingroup

bias and intergroup conflict (Tajfel, 1981; Turner, Hogg, Oakes, Reicher, &Wetherell, 1987).The discussion of identity has had less impact than might be expected,

however, on the quantitative study of political behavior in general and on politicalpsychology more specifically. Despite the recent emergence of identity politicsaround the world, researchers of political behavior have been slow to incorporatethe concept of identity into their empirical studies. This seems odd, given thatdemands for group respect and recognition are at the heart of new social movementsthat argue for the rights of women, religious minorities, diverse ethnic and racial

groups, and gays and lesbians (Taylor, 1994). Such movements cannot be explainedaway as a simple quest for material gain or tangible benefits, and they seem to callfor an explanation that incorporates the notion of identity (see also Monroe, Hankin,& Van Vechten, 2000).

Given their political impact, the emergence of strong social and politicalidentities ought to be of interest to political psychologists, and a theoreticalapproach is needed to advance the study of identity within political science. Oneof the approaches that holds most promise for political psychologists is socialidentity theory, as reflected in the thinking of Henri Tajfel, John Turner, and

colleagues (Tajfel, 1981; Turner, 1996; Turner et al., 1987).1

Social identity theoryis useful for several reasons. It has spawned an enormous number of studies in adiverse group of countries (see Brewer & Brown, 1998). Its key findings, perhapsthe most famous of which is the emergence of ingroup favoritism under the mostminimal of conditions, have been widely replicated (Brewer, 1979; Brown, 1995).It has also generated testable hypotheses that can be applied to a wide range of groups, including those linked to politics. Finally, it addresses the kinds of issuesof interest to political psychologistsintergroup conflict, conformity to groupnorms, the effects of low group status and the conditions under which it generates

collective action, and the factors that promote the categorization of oneself andothers into groups.

Nonetheless, I believe social identity theory has had less impact on politicalpsychology than it might have had otherwise because of various shortcomings andomissions in its research program. In the spirit of constructive dialogue, I criticallyevaluate the utility of social identity theory for political psychology by identifying

1 As will become clear, I focus on social identities and social identity theory but ignore a second strand

of political psychology that has defined identity as more truly individual, something about whopersons are in a deep psychological sense (Young, 1997, p. 32). Inspired by thework of developmentalpsychologist Erik Erikson, political psychologists working in this tradition have investigated, forexample, the psychology of individual leaders, the mindset of altruists, and the psychologicaldevelopment of terrorists (Crenshaw, 1986; Monroe, 1994; Monroe et al., 2000).

128 Huddy

8/9/2019 Huddy

3/30

several key issues that hinder its application to political phenomena. I use thiscritique to outline a research agenda on the nature and impact of identity that cutsacross political and social psychology. I begin with a brief summary of political

research that has incorporated social identity theory, or notions of identity moregenerally, into research on intergroup relations. This is followed by a brief over-view of social identity theory. I then explore in greater detail the challenges posedby political research for social identity theory. Throughout, I argue that socialidentity theorists disinclination to examine the sources of social identity in a realworld complicated by history and culture has placed serious limits on the theorysapplication to political psychology.

Current Research on Political Identity

There are several strands of research in political psychology that have incor-porated the notion of identity. One research strand has emerged around questionsof national identity, patriotism, and multiculturalism. An example is provided byCitrins and Sears investigations of American identity. They have examined thesubjective meaning of being American and uncovered a consensus that it dependson support for the key American values of equality and individualism. Nonetheless,they have also discovered contested aspects of American identity that concern theneed to believe in God or speak up for ones country in order to be considered a

true American (Citrin, Reingold, & Green, 1990; Citrin, Wong, & Duff, 2000). 2And it is these contentious aspects of American identity that mediate the politicalconsequences of national identity. Individuals who support the less consensual,nativist aspects of American identity (such as being Christian) are more likely tooppose policies designed to benefit new immigrants, view negatively the impactof immigration, and believe it is difficult to become American without adoptingAmerican customs (Citrin et al., 1990; Citrin et al., 2000). Other researchers havealso found that the political effects of patriotism depend on its subjective meaning(Schatz, Staub, & Lavine, 1999).

Sears and Citrin also uncovered substantial evidence that members of diverseethnic and racial groups in the United States identify primarily as American andonly secondarily as members of their ethnic or racial group. This is at odds withthe predictions of social identity theory, which suggests that minority groupmembership should be extremely salient to African Americans, Hispanics, andAsians, thus overwhelming national identity (Citrin et al., 2000; Sears, Citrin,

2 In a California poll, some 40% of respondents thought that believing in God was important in makingsomeone a true American. In a national sample (National Opinion Research Center, 1996), 54% of all respondents felt that being a Christian was important in making someone a true American.

From Social to Political Identity 129

8/9/2019 Huddy

4/30

Vidanage, & Valentino, 1994; Sears & Henry, 1999). 3 The inability of simple groupsalience to account for ethnic identity is reinforced in a study by Gurin, Hurtado,and Peng (1994) on national and ethnic identity among Mexican Americans. They

found that Mexican Americans who regularly come into contact with Anglos, andfor whom Mexican ethnicity is therefore highly salient, are no more likely to holdnational (Mexican) or ethnic identities (e.g., Chicano) than are other MexicanAmericans. This raises important questions for social identity theory about theextent to which the salience of ones ethnic or racial groupthe key ingredient inidentity development for many social identity researchersexplains the emer-gence of ethnic and racial identities.

Taken together, research on ethnic and national identities suggests at a mini-mum that identity formation cannot be simply explained by the salience of a group

designation. Rather, it hints at the first of four key issuesthe subjective meaningof identitiesthat I believe need to be addressed by social identity researchersbefore the theory can be successfully applied to political phenomena. As researchon patriotism demonstrates, American identity does not mean the same thing to allAmericans. And it is the meaning of American identity, not its existence, thatdetermines its political consequences. Yet social identity researchers have tendedto ignore this subjective aspect of identities, paying considerable attention to theexistence of simple group boundaries while ignoring their internal meaning.

In a second, related strand of research on ethnic and racial identities, strong

identities have been found to undercut national unity and promote intolerance andintergroup antipathies. Thus, Sidanius, Feshbach, Levin, and Pratto (1997) foundthat a strong identity as a member of a subordinate group in the United States orIsrael (e.g., African Americans in the United States, Arabs in Israel) results in adiminished sense of patriotism. Likewise, Gibson and Gouws (1999) found thatstrong racial and ethnic identities among South Africans increase their perceivedneed for group solidarity, which in turn produces greater antipathy toward out-groups, increases the perception that such groups pose a threat, and promotesintolerance. These findings build on a large body of work that documents the

importance of subjective group membership in shaping political attitudes andbehavior (Conover, 1988; Miller, Gurin, Gurin, & Malanchuk, 1981).

On the surface, these results appear compatible with social identity theorybecause they suggest that membership in a salient minority results in ingroupidentity and outgroup antipathy. Yet upon closer examination, it is clear that thecrucial ingredient in the development of outgroup antipathy in these studies is theexistence of a strong, internalized subjective identity, not simple group member-ship. Moreover, it is clear that not everyone identifies strongly with their ethnic orracial group. These findings thus raise two additional challenges for social identity

3 In contrast, Sidanius et al. (1997) reported that black students who identify with their race are lesspatriotic than black students whodo not. This finding contradicts Sears and Citrins results and is moreconsistent with the predictions of social identity theory.

130 Huddy

8/9/2019 Huddy

5/30

theory. First, how do we explain an individual group members decision to identifyas a group member? This aspect of choice has typically been ignored by socialidentity researchers whose key experimental paradigmthe minimal intergroup

situationassigns members to groups and simply assumes the uniform develop-ment of group identity. Second, social identity theorists typically regard socialidentity as an all-or-none phenomenon. When the group is salient, group identityis paramount. When group membership is not salient, individual identity domi-nates. But how then do we account for identities of variable strength that persistacross situations? When assessed over time, a wide range of group identitiesdemonstrate remarkable stability in both their nature (e.g., African American) andstrength. I believe it is difficult to adapt social identity theory to political phenom-ena without coming to terms with both issuesidentity choice and gradations in

identity strength.A third strand of research in political psychology has focused on the nature of political identities, including an identification with a major political party or theadoption of an ideological moniker as a term of self-description (Abrams, 1994;Duck, Hogg, & Terry, 1995; Duck, Terry, & Hogg, 1998; Kelly, 1989). Deaux,Reid, Mizrahi, and Ethier (1995) examined the social nature of political identitiessuch as conservative, environmentalist, liberal, pacifist, radical, and socialist,concluding that they would expect predictions from social identity theory to bemost applicable to ethnic, religious, [and] political identities because they are

more collective in nature than other individual aspects of identity (p. 286).My own research on feminist identity provides an example of this approach.I apply social identity theory to the development of feminist identity and examinethe ease with which feminist identity changes in response to information about thesocial and political characteristics of feminists and their opponents (Huddy, 1997b,1998). My findings support the extension of social identity theory to politicalidentities and at the same time challenge the theorys view of identities as highlyfluid. In support of a social identity approach, I find that feminist identity dependson feeling similar to the types of women depicted as feminists, independently of

their beliefs (Huddy, 1998). At the same time, I uncover considerable stability infeminist identity that is at odds with Turner and other social categorizationresearchers view that social identities are highly changeable (Haslam, Turner,Oakes, McGarty, & Hayes, 1992; Hogg, Hardie, & Reynolds, 1995). In essence, Ifind that it is difficult to reverse cultural definitions of a typical feminist and, moreimportant, such culturally established group prototypes create a powerful sourceof identity stability (Huddy, 1997b). The considerable stability evinced by diversepolitical identities, not just feminist identity, provides an important fourth chal-lenge to social identity theory that has previously gone unexplored.


8/9/2019 Huddy

6/30

Social Identity Theory: A Brief Overview

In reality, there are two distinct branches of social identity theory: the version

developed by Tajfel (1981) and Tajfel and Turner (1979), known as social identitytheory, and an offshoot developed by Turner and colleagues, referred to asself-categorization theory (Turner et al., 1987). Both theories acknowledge theorigins of social identity in cognitive and motivational factors, although they placediffering emphasis on them (Hogg, 1996, p. 67). The earliest versions of socialidentity theory developed by Tajfel (1981) and Tajfel and Turner (1979) placedkey emphasis on the psychological motivations that lead a group member toendorse or disavow an existing group membership. Turner et al. (1987) havedescribed this motive as a need among group members to differentiate their own

groups positively from others to achieve a positive social identity (p. 42).In contrast, the self-categorization theory developed by Turner et al. (1987)has concentrated on the cognitive underpinnings of social identity. Self-categori-zation theory built on Tajfels earliest cognitive formulations to further develop thecognitive factors that promote categorization of oneself as a group member. Asnoted by Turner et al. (1987), self-categorization theory is a cognitive elaborationof Tajfels earlier theory that provides an explanation for how individuals come toidentify and act as a group (p. 42).

Categorization and Salience

One of the key insights of both social identity theory and self-categorizationtheory is that principles governing the categorization of everyday objects can beextended to explain the categorization of people, including oneself, into socialgroupings. Although this connection between identity formation and the principlesof categorization is developed most fully within self-categorization theory, theinfluence of categorization research on social identity theory was present from thetheorys very beginning. Tajfels early research and theorizing (1981) began from

a purely cognitive perspective, attempting to explain the perceptual distortions thataccompanied categorization (Tajfel, 1981; Tajfel & Wilkes, 1963; for a review of this early work, see Eiser, 1996).

Tajfel then went on to document the astonishing effects of simple socialcategorization, which are quite well known by now. Blue eyes, a preference for thepaintings of Wasily Kandinsky over those of Paul Klee, and calling some peopledot overestimators and others underestimators were sufficient to produce a prefer-ence for fellow group members and to elicit discrimination against outsiders (Allen& Wilder, 1975; Billig & Tajfel, 1973; Brewer & Silver, 1978; Doise & Sinclair,

1973; Tajfel, Billig, & Bundy, 1971; for a summary, see Brewer, 1979). Theexperimental situation popularized by Tajfel and his followers, in which groupswere designated by nothing other than a common label, became known as theminimal intergroup situation (for a review, see Diehl, 1990). In these studies the

132 Huddy

8/9/2019 Huddy

7/30

8/9/2019 Huddy

8/30

group members (Rosch, 1978). Self-categorization researchers believe that it isones perceived similarity to the prototypic group member that plays a key role inthe formation and development of social identity (Hogg, 1996; Hogg & Hains,

1996; McGarty, Turner, Hogg, David, & Wetherell, 1992; Turner et al., 1987).Self-categorization researchers also hold an extremely labile view of socialidentities that seems driven almost completely by ones immediate perceptualcontext. In a paper on Australian stereotypes of Americans, Turner and colleaguesstated that salient self-categories are . . . intrinsically variable and fluid, not merelybeing passively activated but actively constructed on the spot to reflect thecontemporary properties of self and others (Haslam et al., 1992, p. 5). From theirperspective, identities vary in part because social categories (such as age or gender)vary in salience across situations. Indeed, one of the key tenets of self-categorization

theory is that individuals constantly shift back and forth between an individual anda social identity (Brewer & Weber, 1994; Simon, 1997; Turner et al., 1987).It is not just the salience of existing categories, however, that influences the

lability of social identities, according to self-categorization researchers: Theybelieve that categories themselves change across social settings. According toself-categorization theory, individuals are more likely to think of themselves asmembers of social groups under conditions in which the use of a group labelmaximizes the similarities between oneself and other group members, and height-ens ones differences with outsiders (Turner et al., 1987). Thus, categories and their

prototypes spontaneously emerge and change with the attributes of categoryinsiders and outsiders. Hogg et al. (1995) echoed this position when they noted thatsocial identity is highly dynamic: it is responsive, in both type and content, tointergroup dimensions of immediate social comparative contexts (p. 261). Here,type refers to varying category salience, and content implies a change in the groupprototype. In other words, Turner and colleagues believe that group prototypes varyacross social settings and thus contribute further to identity shifts.

Motivational Influences

Social identity is not solely a cognitive matter, however. As Tajfel recognized,there has to be something more to identity to account for ingroup bias and outgroupdiscrimination in the minimal intergroup situation. To round out social identitytheory, Tajfel added motivation to what began essentially as a cognitive model of intergroup perception and discrimination, describing this addition as his secondgreat idea (Turner, 1996, p. 16). According to Tajfel, a need for positive distinct-iveness drives social identity. This means that group identity is likely to emergeamong members of a high-status group because membership positively distin-

guishes group members from outsiders; in contrast, the development of groupidentity is less certain among members of low-status groups who need to addition-ally develop an identity around alternative, positively valued group attributes

134 Huddy

8/9/2019 Huddy

9/30

(social creativity) or fight to change the groups negative image (social change)before membership can enhance their status (Tajfel & Turner, 1979).

Several strands of research demonstrate the motivational underpinnings of

social identity. Tajfel and Turner (1979) suggested that one option available tomembers of low-status groups, especially groups in which membership is perme-able, is to deny ones group membership or identify with an alternativehigher statusgroup. They referred to this strategy as social mobility, and several researchers haveprovided evidence of its existence among members of low-status groups (Jackson,Sullivan, Harnish, & Hodge, 1996; Taylor, Moghaddam, Gamble, & Zellerer,1987; Wright, Taylor, & Moghaddam, 1990). Members of low-status groups canalso resort to the tactics of social creativity and social change to enhance theirgroups standing. Again, there is evidence to back this up. For instance, ingroup

members tend to elevate the importance of positive ingroup characteristics thatconfer superiority over an outgroup in defining their group (Mummendey &Schreiber, 1984; van Knippenberg, 1978; van Knippenberg & van Oers, 1984).Lalonde (1992) observed this strategy in action among members of a losing hockeyteam who acknowledged that their competitors held superior skills but rated theiropponents more negatively on other dimensions. Jackson et al. (1996) found thatmembers of a negative group attempted to change their groups status by rating anundesirable attribute more positively or rating the group more favorably on othercomparative dimensions.

These findings on positive distinctiveness have their parallel in research onethnic identity in which identity is more strongly developed among members of higher status groups. Huddy and Virtanen (1995) found that national identity ismore strongly developed among Cubans than among other Latinos because theybelieve their social status far exceeds that of Mexican Americans or Puerto Ricans.Ethier and Deaux (1994) showed that Hispanic students in their first year of collegeat an Ivy League university who find the university environment threatening totheir Hispanic identity view their group as having lower status, which in turnweakens their identification as Hispanic. In a similar vein, Swann and Wyer (1997)

found that men are more likely to think of themselves in gender-stereotypictermsand thus identify with their genderwhen in the minority, whereaswomen, members of a lower status group, are not as likely to stereotype themselveswhen in the minority.

Some researchers have reasonably equated the need for positive distinctive-ness with the maintenance of personal self-esteem, and they argue that groupmembers with low self-esteem should be more motivated than others to boost thegroups standing and exhibit ingroup bias. However, a number of studies indicatethat it is people with high, not low, self-esteem who are most likely to derogate an

outgroup in order to protect group standing (Crocker & McGraw, 1984; Crocker,McGraw, Thompson, & Ingerman, 1987; Long & Spears, 1998; for a summary of findings, see Abrams & Hogg, 1988). Recent research that distinguishes betweenpersonal and group esteem suggests that the two need to be separated to understand


8/9/2019 Huddy

10/30

the way in which positive distinctiveness works. There is every reason to think thatindividuals would prefer to be associated with positively esteemed groups, irre-spective of their level of self-esteem (Luhtanen & Crocker, 1992). Indeed, Mullen

et al.s (1992) meta-analysis uncovered significantly higher levels of ingroup biasamong members of high-status groups (although there is not complete unanimityon this point; see Long & Spears, 1998).

More recent research suggests several other motives for the development of ingroup identity and ingroup bias, although these await further empirical verifica-tion. Brewer (1991, 1993) suggested that ingroup identity depends on a balancebetween the need to belong and the need for uniqueness, countervailing motivesthat she combined within optimal distinctiveness theory. According to Brewer,identities need to confer the optimal mix of distinctive and common attributes, thus

explaining why members of majority groups evince weaker ingroup identities thando members of minority groups.There is also continued debate over the role of common fate and outgroup

threat as a determinant of ingroup bias, with some studies continuing to report theemergence of ingroup bias only under conditions of intergroup competition(Brewer, 1979; Insko, Schopler, Kennedy, & Dahl, 1992; Rabbie, Schot, & Visser,1989). For example, Flippen, Hornstein, Siegal, and Weitzman (1996) contrastedthe influence of salience and threat on ingroup bias, and they found that it onlyemerges when group members are threatened by outsiders.

Finally, Mullin and Hogg (1998) have introduced an additional motive toaccount for ingroup bias. They argued that ingroup bias emerges in the minimalintergroup situation because group members feel uncertain about their views andidentify with other group members, especially typical group members, to dispelthis unpleasant feeling. This need for certainty was first hinted at by Tajfel (1969),who suggested that the search for coherence may underlie the development of stereotyping and prejudice.

Challenges to Social Identity Theory

This brief overview indicates that social identity theorists have spent timethinking about both the origins of social identity and the development of ingroupbias. Unfortunately, researchers have not allocated their efforts evenly to these twoissues. Research findings provide ample empirical evidence of the consequencesof group membership for intergroup conflict but shed considerably less light on thedevelopment of identity. This is a serious omission for political behavior re-searchers who are interested not only in what happens once group distinctions aremade salient but also in the development of identities, especially strong identities

that endure across situations and over time. Evidence uncovered by social identityresearchers that simply belonging to a group fuels ingroup bias (in the absence of intergroup competition) is a powerful addition to research on intergroup relations.

136 Huddy

8/9/2019 Huddy

11/30

But political behavior researchers are often struck by the absence of group conflictdespite the existence of distinct and salient groups, or by the weakness of identities(e.g., Asian American) among members of salient groups. This raises a politically

important question: Why, despite salient group distinctions, do individuals vary inthe degree to which they identify with a group? The answer matters if strongidentities motivate group-related action (see below). Each of us has many potentialidentities derived from diverse group memberships, but relatively few of theseidentities develop or become politically consequential.

Social identity theory has been faulted for ignoring the powerful identities thatcreate the kinds of intergroup conflicts of interest to political psychologists.Nonetheless, I would like to make a case for studying identities that vary on acontinuum from weak to strong. It is important to understand what turns a weak or

nonexistent identity into something that can motivate ethnic hatred. But this processwould be difficult to understand if all we examined were the very weak identitiesthat arise in the minimal intergroup situation, or the very powerful identities thatcharacterize ethnic or national conflicts. Iris Marion Young (1997) provided anexample of this strengthening process in her discussion of women as a socialcollective. She drew on an incident in a novel by Meredith Tax to describe thetransformation of a group of Russian Jewish immigrant women, on the lower eastside of Manhattan a century ago, from women who shopped at the same butcher toa collective that organized a store boycott to protest local chicken prices. In this

example, a weak former identity is strengthened to the point where it motivatescollective action. The process underlying this transformation deserves greaterscrutiny than it has received so far.

Acquired Versus Ascribed Identities

The first challenge that confronts a politically relevant social identity theoryis to account for the existence of identities acquired by choice. The historicaldevelopment of identity from something ascribed by others to something acquired

by oneself has been discussed with great erudition by political theorist CharlesTaylor (1989) and psychologist Roy Baumeister (1986). Both alluded to the shiftin modern identity from attributes that were essentially determined at birth inmedieval timesones religion, occupation, and economic status in lifeto iden-tities that are much less deterministic and more subject to choice in the modern era.Religion, education, occupation, sexual preference, and domestic roles can now befashioned at will to a much greater extent than was possible in the past (Giddens,1991). This ability to recreate and refashion ones identity many times over isarguably at its extreme in contemporary American society, characterized by its high

levels of residential mobility, second careers, and high divorce rates. As a Polishimmigrant to the United States says about her American acquaintances, everyoneis always on the move and undergoing enormous changes, so they lose track of


8/9/2019 Huddy

12/30

who theyve been and have to keep tabs on who theyre becoming (Sarup, 1996,p. 5). This refashioning of identity goes hand-in-hand with the modern desire forauthenticity and external recognitionfinding ones true self and having it ac-

knowledged by others (Taylor, 1994).The importance of individual choice in identity acquisition holds even for whatwe think of as quite fixed characteristics such as race and ethnicity. This is welldocumented by Nagel (1995), who examined the increasing number of people whoreport an American Indian race in the U.S. Census. Between 1960 and 1990, thenumber of people identifying as American Indian more than tripled, from just over500,000 to almost 2 million. As noted by Nagel, this cannot be explained solely byan increasing birth rate; it also reflects ethnic switching. Nagel found the greatestincrease in American Indian identity among individuals living in urban areas or

non-Indian states without reservations, who have intermarried, speak Englishexclusively, and do not assign their children an Indian race. In other words,American Indian identity has increased among those individuals residing in partsof the country that permit a wide range of ethnic options (p. 953). The ability toacquire (or lose) American Indian identity is less prevalent among Native Ameri-cans living on reservations, for example, who are assigned a race or ethnicity forofficial administrative purposes.

The existence of acquired identities thus poses a crucial challenge for socialidentity researchers. As should be clear from the preceding review of social identity

theory, the minimal intergroup situation on which so many social identity studiesdepend simply does not allow for identity choice. In these studies, participants areessentially assigned to groups and are assumed to internalize their group member-ship. Research participants are randomly assigned to be dot overestimators orunderestimators, lovers of the paintings of Klee or Kandinsky. There is no choiceof identity and no exploration of individual differences in the willingness to adoptsuch experimentally ascribed identities [see Perreault & Bourhis (1999) for asimilar criticism of the minimal intergroup situation]. In the extreme, social identitytheory researchers suggest that the salience of ones group membership is the sole

determinant of identity. Gender identity should be paramount for women who work in male-dominated occupations or work settings. African Americans who work inpredominantly white settings should have difficulty thinking of themselves in otherthan racial terms. But this remains a deeply deterministic view of identity devel-opment that omits individual choice. Salience, one of the key forces behind identityshifts (according to social identity researchers), is a feature of situations, notindividuals.

Identity choice matters because it is a common feature of social identitiesoutside the laboratory. But it may also enhance the development of ingroup

cohesion and outgroup discrimination even within a lab setting. In one of the fewsocial identity studies to examine acquired identities, Turner, Hogg, Turner, andSmith (1984) reported a study in which participants were either assigned or couldchoose to belong to one of two teams competing in a problem-solving exercise.

138 Huddy

8/9/2019 Huddy

13/30

Members of winning teams indicated higher self-esteem and cohesion when theyhad been assigned to the team. But members who voluntarily chose their teamswere more likely to report high self-esteem and group cohesion when they had lost,

suggesting a stronger sense of group commitment when identity is acquired thanwhen it is ascribed. Perreault and Bourhis (1999) extended this research to includethe effects of identity acquisition on the development of outgroup discrimination.They found that group identification increases in strength with the sense that labgroup membership is voluntary. Moreover, strong ingroup identification in thisstudy increased discriminatory behavior against an outgroup in a resource alloca-tion task.

Individual Differences in Identity Acquisition

The notion that social identities are more often acquired than ascribed hints atthe importance of individual differences in the process of identity acquisition, anissue that has been largely ignored by social identity researchers. Is there, forinstance, individual variation in the general proclivity to identify with socialgroups? Duckitt (1989) suggested that authoritarian behavior can be explained inpart by the stronger tendency of some individuals to identify with dominant socialgroups, such as whites in the United States or Christians in western Europe. Cansuch tendencies be accounted for by basic personality traits such as an intolerance

of ambiguity, a need for coherence, or the absence of an openness to experience?Perhaps individuals who are less open to experience or intolerant of ambiguityprefer ascribed to acquired identities and feel uncomfortable with the myriadidentity choices that confront individuals in contemporary society.

In one of the few studies to directly examine individual differences in identityacquisition, Perreault and Bourhis (1999) explored the effects of ethnocentrism,authoritarianism, and personal need for structure on strength of ingroup identifica-tion in an experimentally created lab group. They found that all three personalitymeasures are correlated with strength of group identification, but that these rela-

tionships with identification appear to be driven by ethnocentrism. In other words,individuals who express antipathy toward outsiders are more likely to adopt aningroup identity in the lab. In some ways, Perreault and Bourhis findings raisemore questions than they answer. What are the origins of a general dislike of outsiders? Does this drive the desire for an ingroup identity? Or are there additionalunderlying personality attributes that explain both ethnocentrism and the adoptionof ingroup identity? Obviously, more research is needed to untangle the personalitytraits most likely to influence the adoption of group identity. Other individualdifferences that deserve consideration include the motivational factors discussed

above as possible determinants of ingroup biashigh self-esteem, the need tobelong, the need for uniqueness, and the need for certainty. The study by Perreaultand Bourhis is an encouraging first step in this direction. It is extremely importantfor political psychologists to understand why some individuals in a given social


8/9/2019 Huddy

14/30

and political context adopt a group identity, whereas others in identical circum-stances do not.

Group Differences in the Freedom to Acquire Identity

Groups also differ in the extent to which they allow individuals the freedomto acquire or discard a group identity. Both a groups permeability and the degreeof ambiguity surrounding group membership are likely to influence identityacquisition. Some studies have begun to examine the permeability of groupboundaries, especially for groups that differ in social standing, and have found thatgroup members are quite willing to discard membership in a low-status group(Jackson et al., 1996). Indeed, research by Wright (1997) suggests that boundary

permeability need not be very extensive for group members to contemplateindividual rather than collective solutions to problems of low ingroup status. Thisfinding hints at the existence of weak group identities among members of perme-able groups.

Permeability is not just a feature of highly fluid groups; it can also characterizemembership in relatively fixed groups based on ethnic and regional boundaries.Mummendey, Kessler, Klink, and Mielke (1999) found that East Germans differin how easy they think it is for an East German person to be considered WestGerman, and that individuals who think passing as West German is quite difficult

hold stronger East German identities. In contrast, East Germans who view regionalboundaries as relatively more permeable are more likely to adopt West Germanidentity and are, in turn, more likely to think of themselves as just German.

Questions of group permeability raise concomitant questions about the influ-ence of external labeling on identity acquisition. If group membership is obviousto others, it will be more difficult for a group member to avoid being labeled assuch. It may be relatively easy for an East German to pass as someone from theWest, but much more difficult for an African American to avoid being labeledblack. Less permeable group boundaries and a higher incidence of external labeling

should increase the likelihood that a group member will internalize group identity.Relevant external cues include skin color, gender, group-specific facial and otherphysical features, language, and cultural practices, although the latter two areobviously easier to change than overt physical characteristics. Conversely, attrib-utes that can be hidden or disguised enhance the role of choice in identityacquisition (see McKenna & Bargh, 1998).

Groups also vary in the ambiguity of group membership, a related but separatepoint. Social identity theory from its inception has assumed the existence of fixedand known group memberships. And empirical studies have concentrated on

research with unambiguous naturally occurring groups or experimentalgroups withclearly defined boundaries. But membership in some groups does not neatly fit thisprofile. This may be especially true for groups defined on the basis of politicalideology or beliefs. For many people, the boundaries of political groups (with the

140 Huddy

8/9/2019 Huddy

15/30

exception of political clubs or organizations) are probably vague and difficult todiscern. Is someone who endorses legalized abortion and needle exchange pro-grams but also proposes smaller government and free-market principles a liberal?

What is the demarcation point between liberal and moderate? At what point doesmoderate shade into conservative? These questions are difficult to answer, and theymuddy popular understanding of ideological labels.

The boundaries of political categories are more vague than social categoriesbased on ethnicity or race, for example. Although a person of mixed-race parentagemight ponder whether to describe herself as black or Latino, there is no questionthat she can legitimately claim membership in one or both groups. But this is notthe case for political categories. Thus, although it is possible to paraphrase Tajfeland define identification with various sociodemographic groups based on age, race,

or ethnicity as a self awareness of ones objective membership in the group and a psychological sense of attachment to the group (Conover, 1984, p. 761), thisdefinition is more difficult to apply to groups whose membership criteria remainambiguous. The impact of ambiguous group membership on identity acquisitionand retention has received much less attention than the permeability of groupboundaries, but it would not be surprising to find that it too inhibits the adoptionof group identity, especially when group membership holds negative connotations.

Boundary Versus Meaning

The existence of acquired identities leads to further questions about the basisfor such identity choices. To better understand how identities are acquired, it ishelpful to consider the distinction that has emerged in several different lines of research between belonging to a social category and internalizing its meaning.Anthropologist Frederick Barth (1969, 1981) called this the difference betweennominal identity based on a name and virtual membership based on an experience.Others have referred to this as the difference between a category in which individu-als are united by some common characteristic apparent to outsiders and a group

in which members are aware of their similarities and define themselves on thatbasis (Jenkins, 1996, p. 23). Young (1990) construed this as the difference betweena superficial association in which individuals retain their sense of individualidentity and a group that constitutes part of the individual self. In her view,membership in an association is equivalent to adding another adjective to onesself-descriptionanalogous to acknowledgment of a common group bound-arybut conveys little more about shared experiences or a common outlook. Onthe other hand, membership in a group shapes and influences an individualsidentity.

For me, this distinction embodies the difference between group boundariesand the meaning of group membership. As already noted, social identity researchhas focused on the extent to which group boundaries define group membership and


8/9/2019 Huddy

16/30

shape the adoption of group identity. But this emphasis on boundaries has occurredat the expense of the meaning of group membership (Deaux, 1993). If all groupidentities were simply ascribed to group members, a knowledge of group bounda-

ries may be all that is needed to understand the consequences of group membership.But when group identities are acquired, the meaning of group membership mayhave a powerful influence over the voluntary adoption of identity and its conse-quences once acquired.

It is difficult for political psychologists to focus exclusively on group bounda-ries because the very process of labeling groups, and thus defining group bounda-ries,becomes ensnared with the meaning of group membership. Consider the labelsfor ethnic and racial groups in the United States. It is sobering to discover that theU.S. census has used a different set of categories for racial and ethnic groups in

every national census (Martin, DeMaio, & Campanelli, 1990). The census classi-fication of Mexican Americans affords an interesting example of this inconsis-tency. In 1930 Mexicans were counted as non-white; in 1940 they were consideredas persons of Spanish mother tongue; in 1950 and 1960 they were regarded as whitepersons of Spanish surname, and in 1970 as persons of both Spanish surname andSpanish mother tongue (Fox, 1996). Disputes over the meaning of group member-ship also result in battles over who draws and defines group boundaries. Theadoption of the term African American, championed by Jesse Jackson, carries withit notions of African ancestry that alter the meaning of black identity and may not

appeal to all, or even many, black Americans (Martin, 1991). Feminists in theUnited States battled among themselves in the late 1960s and early 1970s over whocould and who could not be rightfully considered a feminist, with conflicts eruptingover ones political ideology, sexual preference, and the gender of ones children(Ryan, 1992).

An emphasis in social identity research on groups that lack meaning mayseriously hamper our understanding of both identity acquisition and its conse-quences. Consider German nationalism. For obvious reasons, many Germans feelsome lingering unease at the notion of strong German nationalism and resist a

patriotic identity even when their German identity is made salient. As empiricalevidence, Schwartz, Struch, and Bilsky (1990) found that German students do notexpect other Germans to evince ingroup bias against Israelis in a resource allocationtask, but Israeli students predict the emergence of ingroup bias among IsraelisagainstGermans. Quite clearly, suchexpectations arise from the history of relationsbetween Germans and Jews, not the salience of their respective national identities.

In diverse groups, group members may attach different meanings to groupidentity (Cohen, 1986; Jenkins, 1996). Diverse meanings arise when the samegroup exists in different regions of a country or when the same group emerges

among distinct national subgroups or subcultures. It can also occur when themeaning of group membership is contested, perhaps for political reasons. Suchdifferences in meaning can have a dramatic impact on the consequences of identity,as seen above in research by Citrin and colleagues on American identity. Mexican

142 Huddy

8/9/2019 Huddy

17/30

identity in the United States affords an example of this phenomenon. MexicanAmericans born in the United States hold a binational identity as both Americanand Mexican American (Gurin et al., 1994). Individuals born in Mexico, however,

associate being Mexican American with being Mexican and rarely think of them-selves as just American. In addition, being Latino or Hispanic is tied to a politicized,pan-Hispanic identity for U.S.-born but not for Mexican-born individuals. Obvi-ously, such findings suggest that it would be more difficult to unite nonU.S.-bornMexican Americans around the terms Hispanic or Latino, which they do not regardas inherently political.

To complicate matters further, the internal meaning of a group can be quitedifferent from its meaning to outsiders (Cohen, 1986). Group members attemptsto elevate their groups standing and redefine negative identities play a role in this

discrepancy. Group members may even choose to internalize a group identitybecause their conception of what group membership means is different from thatof potential members who fail to adopt the identity. The important point is that weneed to examine the conception of both insiders and outsiders to arrive at themeaning of group membership.

Of course, it is relatively easy to urge the further study of group meaning, butquite another matter to actually do so. Tragically, there is no shortcut. Meaning iscreated over time by culture and history and requires careful investigation. Thissounds daunting, yet social identity theory suggests several key places to begin the

search. I briefly consider four factors that help to imbue group membership withmeaning: the valence of group membership, the defining social characteristics of typical group members, the core values associated with membership, and thecharacteristics of common outgroups who help to define what the ingroup is not.

Valence of group membership. As summarized in the earlier overview of social identity theory, identity development seems to be inhibited among groupsthat are viewed negatively, especially when group boundaries are permeable. 4

Individual differences in the perception of a groups valence may also help toaccount for identity development. Some Americans, for example, are proud of their

overseas image and look forward to encountering fellow Americans outside theUnited States; others cringe when they hear an American accent in a Paris caf andattempt to flee the scene as quickly as possible in anticipation of local anti-Ameri-canism. It may be possible to transform such reluctant Americans into patriotsunder special circumstances, but it is certainly more difficult than for Americanswho willingly embrace American identity. Moreover, the valence of identitydepends on the meaning attached to the symbols of patriotism. Not surprisingly,

4 There is some tension between the influence of low group status and group salience on identityacquisition. To the extent that low group status heightens group salience, it may actually enhanceidentity acquisition. But this effect needs to be distinguished from the effects of low group statusindependent of group salience, whose effects are quite opposite and serve to hinder the developmentof group identity.


8/9/2019 Huddy

18/30

Americans who came of age during Vietnam feel less patriotic and are less attachedto symbols such as the flag than are older Americans of the World War IIgeneration. These examples show the potential of studying the valence of group

membership among those who do and do not identify with the group, and amongthose who can and cannot be reasonably considered potential group members. Identification with a prototype. Self-categorization researchers have high-

lighted the importance of a group prototype or typical group member in defininggroup membership. The prototype approach suggests that greater attention shouldbe paid to the types of people who typically exemplify group membership (andgive it meaning). According to self-categorization theory, group members simi-larity to the group prototype should enhance identity development. The definingcharacteristics of the prototype may also hold the key to understanding group

members behavior. An in-depth analysis of a group prototype should help touncover the existing basis of similarity that drives group identity and the kinds of people who are most and least likely to adopt group identity. If conservativesouthern male congressional representatives in the United States exemplify thecontemporary Republican, it would not be surprising to find that working womenin the northeastern and western United States distance themselves from the Repub-lican party. If Tony Blair is synonymous with the Labor party in Britain, blue-collarunion members may be reluctant to identify as Labor party supporters. The newsmedia are an especially good place to begin the search for the characteristics of

prototypes that exemplify social and political groups, given the narrow range of people who appear in the news (Huddy, 1997a). Such objective analysis shouldbe coupled with the subjective impressions of the group prototype among membersof both the ingroup and the outgroup.

Core values. In addition to the characteristics of typical group mem-bersovert signals that can be expressed in dress, language, and lifestylemean-ing can also be gleaned, according to Barth, via group members basic valueorientations. Fox (1996) provided a fascinating account of the effort to forge apan-Hispanic identity in the United States through an emphasis on common values.

In his view, one of the unifying themes that has emerged to describe Hispaniccommonality is a shared support for populist democracy, with an emphasis onpersonal liberty and support for the little guy. Indeed, Fox gave political valuesa more central role in the creation of a pan-Hispanic identity than shared history,a common language, or similar ethnic background.

Research by Schwartz et al. (1990) illustrates one way to assess the valuesunderlying group membership. In their study of German and Israeli studentsreferred to earlier, students ranked 19 terminal and 18 instrumental values on thebasis of their own preference order and that of their national group. Not surpris-

ingly, ones own views and those of ones group are related, although this link isstronger for Israeli than for German students. This suggests that an important sourceof national identityshared valuesis stronger among Israeli than among German

144 Huddy

8/9/2019 Huddy

19/30

students and hints at an important source of weakened national identity amongGermans.

Differences from outgroups. Outgroups do more than signal group bounda-

ries; they also communicate information about what the group is not. This notionis integral to the view of categories advanced by Lakoff, Rosch, and others and hasbeen fully incorporated into the thinking of self-categorization researchers. Yetalmost no one has examined the portrayal of outgroup members to shed light onthe meaning of category membership. What happens, for instance, when anappealing group emerges as the enemy? During the battle over the Equal RightsAmendment, Phyllis Schlafly helped to define the meaning of feminism for manywomen by demonstrating that homemakers and women who were not pursuingcareers were outside the feminist label. The notion that outgroups help to define

category membership is linked to Barths (1981) view that much of the meaningof identity is created at its boundaries in interaction or dialogue with outgroupmembers. One obvious political implication of this finding is that group identitymay be more diffuse and less intense in the absence of a clear outgroup to sharpenthe meaning of group membership and identify the kinds of people who lie outsidethe group boundary.

Shades of Group Identity

There is growing recognition among identity researchers that the effects of group membership depend to some degree on identity strength. This evidence isoften interpreted as consistent with social identity theory, although I perceiveinconsistencies here between an emphasis on identity strength and current thinkingamong social identity researchers. Mullin and Hogg (1998) provided an exampleof some of these inconsistencies. They acknowledged, for instance, that intergroupdiscrimination depends in part on the degree of ingroup identification, but wenton to discuss how group identity results in a depersonalization or a fusing withother group members that tends to reduce ones sense of individuality, minimize

ingroup differences, and promote conformity to the group prototype. The loss of individual identity that accompanies the emergence of group identity sounds likean all-or-nothing phenomenon that does not easily accommodate shades of groupidentity.

I am concerned that the identity continuum advanced by social identityresearchers, anchored at one end by social identities and at the other by aspects of individual identity, leads to a very stark view of identity that is at odds with reality.As we know from survey research, social identities are adopted by degrees andrepresent something intermediate between an all-encompassing group identity and

a distinctively unique persona. This is certainly true for political identities. If anything, younger Americans demonstrate an increasing aversion to extremepolitical identities when asked to indicate whether they think of themselves as


8/9/2019 Huddy

20/30

strong or not-so-strong Democrats or Republicans, liberals or conservatives(Abramson, 1976, 1979; Keith et al., 1992). The labels independent and mod-erate have increased in popularity while the number of strong identifiers has

declined over the last several decades. In a similar vein, women are more likely tocall themselves feminists if they can qualify feminist identity by indicating thatthey are not especially strong feminists (Huddy, Neely, & LaFay, 2000).

I believe these shades of identity arise from feeling closer to or farther awayfrom a group prototype or key values endorsed by prototypic members. Theexistence of shaded identities only becomes clear once we move beyond a view of social identity as inclusion inside a group boundary to look more closely at theinfluence of meaning on identity development. Boundaries connote an all-or-noth-ing membership; meaning holds out the possibility of degrees of similarity. Forcing

group members to think of themselves either as group members or as distinctindividuals misses the complex nature of identity, which is simultaneously indi-vidual and social (Deaux, 1993; Jenkins, 1996).

More important, there is evidence from Branscombe and her colleagues (Noel,Wann, & Branscombe, 1995; Wann & Branscombe, 1990, 1993) that shades of group identity influence the development of ingroup bias and outgroup derogation. 5

Other studies replicate their findings. Japanese students who identify strongly withtheir vocational school are less likely than weak identifiers to denigrate their fellowingroup members after reading negative information about their group (Karasawa,

1991). Perreault and Bourhis (1999) found that individuals who identify morestrongly with their group are more likely to discriminate against an outgroup in aresource allocation task. Germans with stronger regional identity feel more posi-tively about their region than about the nation and demonstrate greater regionalhomogeneity (Simon, Kulla, & Zobel, 1995).Purdue students who identify stronglywith their school exhibit higher levels of ingroup bias and ingroup pride (Jackson& Smith, 1999). Individuals with a strong identity as a member of a marginalizedgroup (e.g., sexual and political) are more likely than those with a weak identity toaccept their identity, share it with friends and family, and feel less estranged from

society when they participate in a group-related electronic news group (McKenna& Bargh, 1998).

There is even some suggestion that the strongest forms of identity may be theleast affected by context. Kinket and Verkuyten (1997) differentiated strength of ethnic identity among Turkish and Dutch schoolchildren aged 10 to 13 who attendprimary school in the Netherlands. They distinguished ethnic self-identification(In terms of ethnic group, I consider myself to be . . .) and self-description (usinga measure analogous to Kuhn and McPartlands Twenty Statements Test) fromethnic self-evaluation [using aspects of Luhtanen & Crockers (1992) self-esteem

scale; e.g., I feel good about being Turkish] and the introjection of ones ethnic

5 For an exception, see Hinkle and Brown (1990).

146 Huddy

8/9/2019 Huddy

21/30

group (If someone said something bad about Turkish people, would you feelalmost as if they had said something about you?). They found that the highest (orstrongest) level of identity (introjection) is unaffected by classroom context (e.g.,

percentage of Dutch and Turkish students), whereas the lowest (or weakest) levelis most affected. As the authors noted, these findings suggest that social identitytheory may have actually overestimated the effects of situational salience onidentity by focusing on relatively weak identities created in the lab.

Political psychologists have always included measures of identity strength intheir research, and this measurement approach finds vindication in the results of the recent social psychological studies reviewed here. But more needs to be doneto incorporate the notion of identity strength into social identity theory, especiallythe transition from weak to strong identity. Identity strength can be assessed using

the traditional approach, which asks whether one is a strong or not-so-strongidentifier. But new approaches are also emerging in social psychological research.One of the most interesting of these is based on work by Aron and colleagues (Aron,Aron, & Smollan, 1992; Aron, Aron, Tudor, & Nelson, 1991). Smith and Henry(1996) developed Arons method to assess the extent to which automatic attitudesabout a groups attributes are incorporated into the self-concept. Group memberswho respond more rapidly to traits that are characteristic of both themselves andan ingroup are assumed to have internalized their group identity more completely.This approach may prove to be an important tool for examining individual differ-

ences in identity strength and may provide deeper insight into the processof identitydevelopment.

Identity Stability

There is continued disagreement among researchers on the relative stabilityand fluidity of social and political identities. On the one hand, social identityresearchers tend to emphasize the fluidity of identity, highlighting how identitieschange with social context. On the other hand, social identities such as partisan and

ethnic identity demonstrate remarkable stability over time when assessed in sur-veys on social and political topics, and they are much more stable than a range of other social and political attitudes (Alwin, Cohen, & Newcomb, 1990; Converse& Markus, 1979; Ethier & Deaux, 1994; Sears, 1983; Sears & Henry, 1999). Thediscrepancy between social identity researchers view of identities as highly fluidand the remarkable stability of social and political identities observed in panelstudies needs to be resolved.

Questions about the relative stability of social identities hold particular interestfor political scientists. A highly fluid and contingent view of identity clashes with

the political reality of newly emergent independence and social movements aroundthe world that argue for the rights of women, diverse ethnic and racial groups, andgays and lesbians. The sustained commitment that underlies the actions of individuals


8/9/2019 Huddy

22/30

in such movements seems at odds with the notion of identities as highly contingentand changeable. On the other hand, there is abundant evidence from everydaypolitics that political and national identities can be manipulated by the words and

actions of political leaders, can shift in intensity with normative support for amovements goals, and can vary in salience across settings.John Turner and colleagues are some of the strongest proponents of the view

that social identities are highly labile. Hogg and Turner (1985) found, for example,that increasing the salience of study participants gender increases the likelihoodthat they think of themselves in gender-stereotypic terms. But, as noted above,self-categorization researchers also believe that categories themselves changeacross social settings. This spontaneous emergence of social categories revolvesaround the temporary formation of a group prototype or typical group member who

embodies the groups distinctive attributesthose that are shared among groupmembers but are absent among non-members.Yet the views of Turner and colleagues do not capture the substantial stability

observed across a range of social and political identities. As I have noted, onepersistent criticism leveled at social identity theory is that much of its empiricalbase depends on information about identities that are relatively weak or nonexistentprior to the experimental setting in which they are created. It is hard to believe thatlongstanding political identities linked to major ideologies or political partieswould exhibit the same high level of fluidity as an artificial identity created in the

lab. Kinket and Verkuytens (1997) evidence that strong identities are moreresistant to social context, discussed above, supports this point.It is also difficult to believe that a group prototype can be changed as easily as

suggested by self-categorization researchers. As we know from numerous stereo-type studies, there are many ways in which respondents can rationalize theexistence of an exceptional group member without shifting their underlying imageof the group as a whole. One explanation for this, provided by Stangor andMcMillan (1992), is that information about unusual group members is discountedbecause groups are expected to exhibit internal diversity. A single exception does

not violate the general rule. In contrast, greater attention is paid to inconsistentinformation about an individual who is expected to demonstrate consistent behav-ior. On the basis of these findings, group prototypes should also be resistant tochange, especially within groups that have an established history and, thus, mean-ing outside a lab setting.

The actual fluidity of group prototypes is difficult to gauge from currentresearch. There is a tendency among self-categorization researchers to simplyassume that group prototypes vary with social context. A study by McGarty et al.(1992) demonstrates this approach. In their research, they construct small groups

of three to five individuals, assess their views on a range of topics, and designatethe prototype as the individual whose views are most like those of other groupmembers and least like the views of outgroup members. In other words, they imposea group prototype on group members in this lab situation. But this is not the same

148 Huddy

8/9/2019 Huddy

23/30

as examining the prototype of a group such as American Jews. It is easy to portraya Woody Allen double as a typical American Jew but harder to fit atypical Jewssuch as the Three Stooges into the category. Does the category prototype change

simply because one is in a situation in which atypical group members predominate?This seems unlikely and raises serious questions about the degree to whichprototypes change with situational factors, and whether social identities are as fluidas suggested by self-categorization researchers.

The two major sources of identity instability in self-categorization re-searchgroup salience and changes in the group prototypedeserve much closerattention than they have received from researchers to date. My research on feministidentity suggests that it is much easier to alter the salience of feminist identity thanit is to shift the nature of the group prototype (Huddy, 1997b). When the word

feminist is included in an experimental news story about the womens movement,it enhances feminist identity among women who like feminists and dampensidentity among those who oppose feminists. These effects are in line with thepredicted effects of group salience. But changing the meaning of feminist identityproves to be more difficult. Altering the description of feminists in the news storyfrom the leaders of a womens rights groupthe typical feministto ordinarywomen, such as homemakers and clerical workers, proves unconvincing to studyparticipants. Women who hold views similar to those expressed by feminists in thestory only adopt feminist identity when such views are expressed by the leader of

a womens rights group. Holding views similar to those of a feminist depicted asan ordinary woman has no effect on feminist identity. Apparently, ordinaryhomemakers, working women, and working mothers stand too far outside thefeminist prototype to realistically convey information about the views of feminists;simply describing these women as feminists does nothing to change this fact.

A Research Agenda

The preceding review and discussion highlights several important directions

for future research on political and social identities. First, it is important to expandthe scope of social identity research to include a range of real-world identities of varied strength. Social identity researchers are paying greater attention to identitiesthat exist outside the lab, but the choice of such identities (e.g., gender, occupation,college major) is rarely discussed explicitly. Greater thought should be given to thechoice of these real-world identities and greaterattention paid to identities that varyin strength. In this proposed research agenda, there is a role for the weak identitiescreated in the lab. But such weak identities cannot form the entire basis for thisresearch. Identities created in the lab exhibit considerable fluidity, yet the evidence

reviewed here suggests that strong identities are resistant to the effects of context.The weak identities studied in the lab may also underestimate the role played byidentity in shaping intergroup conflict. There is no question that identities created


8/9/2019 Huddy

24/30

in the lab result in pervasive ingroup bias. But the effects of strong identities playan even more powerful role in motivating outgroup discrimination.

Second, we need to know more about the interrelated processes of identity

formation and development. It is important to understand how identities areacquired; it is equally important to understand their progression from weak tostrong. Membership in real-world groups is probably on average weaker, andoccasionally much stronger, than the identities observed in a typical social identitystudy. In the real world, weak identities are a product of commonly ambiguouscriteria for group membership and frequently permeable group boundaries. Suchreal-world identities are frequently weaker than those observed in the lab, wheregroup membership is clear-cut and highly salient. On the other hand, somereal-world groups have a well-established group prototype, have a clear-cut enemy,

and are associated with values that are linked to definitive historical moments orcultural practices. This should result in much stronger group identities than thosetypically found in a lab setting. The role of political actors and events in the processof identity crystallization is of particular interest to political psychologists. To whatextent can politicians redirect or intensify identity by making salient specificmeanings of group membership or focusing on a particular enemy? The psycho-logical origins of identity development deserve much greater attention than theyhave received to date.

Third, to better understand the process of identity development, more research

is needed on the characteristics of individuals that predispose them to adopt groupidentity. Some individuals may be very willing to adopt multiple identities, whereasothers prefer to define themselves on the basis of one or two key group affiliations.What are the key personality characteristics that identify such individual differ-ences? Are there some individuals who simply avoid group categorizations alto-gether, preferring to view themselves and others as unique individuals?And at whatpoint are such resistant individuals caught up in societal forces that impel them tointernalize ingroup membership and develop antipathy toward an outgroup?

Fourth, individual differences alone will never completely explain identity

development. If identities were a stable feature of individuals, it would be difficultto account for dramatic changes over time in levels and strength of national andregional identities. To illuminate the process of identity development, we also needto understand the meaning of group identity. Understanding the connotations of group membership for group identifiers, potential identifiers, and outsiders willhelp to explicate the process of identity development. The meaning of a groupidentity may also shed light on the differing consequences of group identity forgroup members behavior, attitudes, and values.

150 Huddy

8/9/2019 Huddy

25/30

8/9/2019 Huddy

26/30

Brown, R. (1995). Prejudice: Its social psychology . Oxford: Blackwell.

Butler, J. (1990). Gender trouble . New York: Routledge.

Citrin, J., Reingold, B., & Green, D. P. (1990). American identity and the politics of ethnic change. Journal of Politics, 52, 11241154.

Citrin, J., Wong, C., & Duff, B. (2000). The meaning of American national identity: Patterns of ethnicconflict and consensus. In R. Ashmore, L. Jussim, & D. Wilder (Eds.), Social identity, inter-groupconflict and conflict resolution . New York: Oxford University Press.

Cohen, A. P. (1986). Belonging: The experience of culture. In A. P. Cohen (Ed.), Symbolisingboundaries: Identity and diversity in British cultures (pp. 117) . Manchester: ManchesterUniversity Press.

Conover, P. J. (1984). The influence of group identification on political perception and evaluation. Journal of Politics, 46, 760785.

Conover,P. J. (1988). Therole of socialgroups in political thinking. British Journal of Political Science,18, 5176.

Converse, P. E., & Markus, G. B. (1979). Plus ca change . . . : The new CPS election study panel. American Political Science Review, 73, 249.

Crenshaw, M. (1986). The psychology of political terrorism. In M. Hermann (Ed.), Political psychology(pp. 379413). San Francisco: Jossey-Bass.

Crocker, J., & McGraw, K. M. (1984). Whats good for the goose is not good for the gander. American Behavioral Scientist, 27, 357369.

Crocker, J., McGraw, K. M., Thompson, L. L., & Ingerman, C. (1987). Downward comparison,prejudice, and evaluation of others: Effects of self-esteem and threat. Journal of Personality and Social Psychology, 52, 907916.

Deaux, K. (1993). Reconstructing social identity. Personality and Social Psychology Bulletin, 19, 412.

Deaux, K., Reid, A., Mizrahi, K., & Ethier, K. A. (1995). Parameters of social identity. Journal of Personality and Social Psychology, 68, 280291.

Diehl, M. (1990). The minimal group paradigm: Theoretical explanations and empirical findings. European Review of Social Psychology, 1, 263292.

Doise, W., & Sinclair, A. (1973). The categorization process in intergroup relations. European Journalof Social Psychology, 3, 145157.

Duck, J. M., Hogg, M. A., & Terry, D. J. (1995). Me, us and them: Political identification and thethird-personeffectin the1993Australian federal election. EuropeanJournalof Social Psychology,25, 195215.

Duck, J. M., Terry, D. J., & Hogg, M. A. (1998). Perceptions of a media campaign: The role of socialidentity and the changing intergroup context. Personality and Social Psychology Bulletin, 24,

316.Duckitt, J. (1989). Authoritarianism and group identification: A new view of an old construct. Political

Psychology, 10, 6384.

Eiser, J. R. (1996). Accentuation revisited. In W. P. Robinson (Ed.), Social groups and identities: Developing the legacy of Henri Tajfel (pp. 121142) . Oxford: Butterworth-Heinemann.

Ethier, K. A., & Deaux, K. (1994). Negotiating social identity when contexts change: Maintainingidentification and responding to threat. Journal of Personality and Social Psychology, 67,243251.

Flippen, A. R., Hornstein, H. A., Siegal, W. E., & Weitzman, E. A. (1996). A comparison of similarityand interdependence as triggers for in-group formation. Personality and Social Psychology

Bulletin, 22, 882893.

Fox, G. (1996). Hispanic nation: Culture, politics and the constructing of identity . Tucson, AZ:University of Arizona Press.

Gergen, K. J. (1971). The concept of self . New York: Holt, Rinehart and Winston.

152 Huddy

8/9/2019 Huddy

27/30

Gibson, J. L., & Gouws, A. (1998). Social identity theory and political intolerance in South Africa.Unpublished manuscript, University of Houston.

Giddens, A. (1991). Modernity and self-identity: Self and society in the late modern age . Cambridge:Polity.

Gurin, P., Hurtado, A., & Peng, T. (1994). Group contacts and ethnicity in the social identities of Mexicanos and Chicanos. Personality and Social Psychology Bulletin, 20, 521532.

Haslam, S. A., Turner, J. C., Oakes, P. J., McGarty, C., & Hayes, B. K. (1992). Context-dependentvariation in social stereotyping 1: The effects of intergroup relations as mediated by social changeand frame of reference. European Journal of Social Psychology, 22, 320.

Hinkle, S., & Brown, R. (1990). Intergroup comparisons and social identity: Some links and lacunae.In D. Abrams & M. A. Hogg (Eds.), Social identity theory: Constructive and critical advances(pp. 4870). New York: Springer-Verlag.

Hogg, M. A. (1996). Intragroup processes, group structure, and social identity. In W. P. Robinson (Ed.),Social groups and identities: Developing the legacy of Henri Tajfel (pp. 6593). Oxford:Butterworth-Heinemann.

Hogg, M. A., & Hains, S. C. (1996). Intergroup relations and group solidarity: Effects of groupidentification and social beliefs on depersonalized attraction. Journal of Personality and SocialPsychology, 70, 295309.

Hogg, M. A., Hardie, E. A., & Reynolds, K. J. (1995). Prototypical similarity, self-categorization, anddepersonalized attraction: A perspective on group cohesiveness. European Journal of SocialPsychology, 25, 159177.

Hogg, M. A., Terry, D. J., & White, K. M. (1995). A tale of two theories: A critical comparison of identity theory with social learning theory. Social Psychology Quarterly, 58, 255269.

Hogg, M. A., & Turner, J. C. (1985). Interpersonal attraction, social identification, and psychologicalgroup formation. European Journal of Social Psychology, 15, 5166.

Huddy, L. (1997a). Feminists and feminism in the news. In P. Norris (Ed.), Women, the media, and politics (pp. 183204). New York: Oxford University Press.

Huddy, L. (1997b, September). Political identification as social identity. Paper presented at the annualmeeting of the American Political Science Association, Washington, DC.

Huddy, L. (1998, September). The social nature of political identity: Feminist image and feminist identity. Paper presented at the annual meeting of the American Political Science Association,Boston.

Huddy, L., Neely, F., & LaFay, M. (2000). Attitudes towards the womens movement. Public OpinionQuarterly .

Huddy, L., & Virtanen, S. (1995). Subgroup differentiation and subgroup bias among Latinos as afunction of familiarity and positive distinctiveness. Journal of Personality and Social Psychology,68, 97108.

Insko, C. A., Schopler, J., Kennedy, J. F., & Dahl, K. R. (1992). Individual-group discontinuity fromthe differing perspectives of Campbells Realistic Group Conflict Theory and Tajfel and TurnersSocial Identity Theory. Social Psychology Quarterly, 55, 272291.

Jackson, J. W., & Smith, E. R. (1999). Conceptualizing social identity: A new framework and evidencefor the impact of different dimensions. Personality and Social Psychology Bulletin, 25, 120135.

Jackson, L. A., Sullivan, L. A., Harnish, R., & Hodge, C. N. (1996). Achieving positive social identity:Social mobility, social creativity, and permeability of group boundaries. Journal of Personalityand Social Psychology, 70, 241254.

Jenkins, R. (1996). Social identity . London: Routledge.

Karasawa, M. (1991). Toward an assessment of social identity: The structure of group identificationand its effects on in-group evaluations. British Journal of Social Psychology, 30, 293307.

Keith, B., Magleby, D. B., Nelson, C. J., Orr, E., Westyle, M. C., & Wolfinger, R. E. (1992). The mythof the independent voter . Berkeley, CA: University of California Press.


8/9/2019 Huddy

28/30

Kelly, C. (1989). Political identity and perceived intragroup homogeneity. British Journal of SocialPsychology, 28, 239250.

Kinket, B., & Verkuyten, M. (1997). Levels of ethnic self-identification and social context. SocialPsychology Quarterly, 60, 338354.

Lakoff, G. (1987). Women,fire, and dangerous things: What categories revealabout the mind . Chicago:University of Chicago Press.

Lalonde, R. N. (1992). The dynamics of group differentiation in the face of defeat. Personality and Social Psychology Bulletin, 18, 336342.

Long, K. M., & Spears, R. (1998). Opposing effects of personal and collective self-esteem oninterpersonal and intergroup comparisons. European Journal of Social Psychology, 28, 913930.

Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scale: Self-evaluation of ones socialidentity. Personality and Social Psychology Bulletin, 18, 302318.

Markus, H. (1977). Self-schemata andprocessing information about the self. Journal of Personality and Social Psychology, 35, 6378.

Martin, B. L. (1991). From Negro to Black to African American: The power of names and naming.Political Science Quarterly, 106, 83107.

Martin, E., DeMaio, T. J., & Campanelli, P. C. (1990). Context effects for census measures of race andHispanic origin . Public Opinion Quarterly, 54, 551566.

McGarty, C., Turner, J. C., Hogg, M. A., David, B., & Wetherell, M. S. (1992). Group polarization asconformity to the prototypical group member. British Journal of Social Psychology, 31, 120.

McGuire, W. J.,McGuire, C. V.,Child,P.,& Fujioka, T. (1978). Salience of ethnicity in thespontaneousself-concept as a function of ones ethnic distinctiveness in the social environment. Journal of Personality and Social Psychology, 36, 511520.

McGuire, W. J., & Padawer-Singer, A. (1976). Trait salience in the spontaneous self-concept. Journalof Personality and Social Psychology, 33, 743754.

McKenna, K. Y. A., & Bargh, J. A. (1998). Coming out in the age of the Internet: Identitydemarginalization through virtual group participation. Journal of Personality and SocialPsychology, 75, 681694.

Miller, A. H., Gurin, P., Gurin, G., & Malanchuk, O. (1981). Group consciousness and politicalparticipation. American Journal of Political Science, 25, 494511.

Monroe, K. R., with Epperson, C. (1994). But what else could I do? Choice, identity and acognitive-perceptual theory of ethical political behavior. Political Psychology, 15, 201226.

Monroe, K. R., Hankin, J., & Van Vechten, R. B. (2000). The psychological foundations of identitypolitics: A review of the literature. Annual Review of Political Science .

Mullen, B., Brown, R., & Smith, C. (1992). Ingroup bias as a function of salience, relevance, and status:

An integration. European Journal of Social Psychology, 22, 103122.Mullin, B. A., & Hogg, M. A. (1998). Dimensions of subjective uncertainty in social identification and

minimal intergroup discrimination. British Journal of Social Psychology, 37, 345365.

Mummendey, A., Kessler, T., Klink, A., & Mielke, R. (1999). Strategies to cope with negative socialidentity: Predictions by social identity theory and relative deprivation theory. Journal of Personality and Social Psychology, 76, 229245.

Mummendey, A., & Schreiber, H. J. (1984). Different just means better: Some obvious and somehidden pathways to in-group favouritism. British Journal of Social Psychology, 23, 363368.

Nagel, J. (1995). American Indian ethnic renewal: Politics and the resurgence of identity. AmericanSociological Review, 60, 947965.

National Opinion Research Center (1996). General Social Survey. Data available fromwww.umich.edu/gss99/

Neisser, U. (1987). From direct perception to conceptual structure. In U. Neisser (Ed.), Concepts and conceptual development: Ecological and intellectual factors in categorization . Cambridge:Cambridge University Press.

154 Huddy

8/9/2019 Huddy

29/30

Noel, J. G.,Wann, D. L., & Branscombe,N. R. (1995). Peripheral ingroup membership statusandpublicnegativity toward outgroups. Journal of Personality and Social Psychology, 68, 127137.

Novotny, K. M. (1998). Taylor-made? Feminist theory and the politics of identity. Women and Politics,19, 118.

Perreault, S., & Bourhis, R. Y. (1999). Ethnocentrism, social identification, and discrimination.Personality and Social Psychology Bulletin, 25, 92103.

Rabbie, J. M., Schot, J. C., & Visser, L. (1989). Social identity theory: A conceptual and empiricalcritique from the perspective of a behavioural interaction model. European Journal of SocialPsychology, 19, 171202.

Rosch, E. (1978). Principles of categorization. In E. Rosch & B. B. Lloyd (Eds.), Cognition and categorization (pp. 2748) . Hillsdale, NJ: Erlbaum.

Ryan, B. (1992). Feminism and the womens movement: Dynamics of change in social movement ideology and activism . New York: Routledge.

Sarup, M. (1996). Identity, culture and the postmodern world . Edinburgh: Edinburgh University Press.

Schatz, R. T., Staub, E., & Lavine, H. (1999). On the varieties of national attachment: Blind versusconstructive patriotism. Political Psychology, 20, 151174.

Schwartz, S. H., Struch, N., & Bilsky, W. (1990). Values and intergroup social motives: A study of Israeli and German students. Social Psychology Quarterly, 53, 185198.

Sears, D. O. (1983). The persistence of early political predispositions: The roles of attitude object andlife stage. Review of Personality and Social Psychology, 4, 79116.

Sears, D. O., Citrin, J., Vidanage, S., & Valentino, N. (1994, September). What ordinary Americansthink about multiculturalism. Paper presented at the annual meeting of the American PoliticalScience Association, New York City.

Sears, D. O., & Henry, P. J. (1999). Ethnic identity and group threat in American politics. The PoliticalPsychologist, 4 (2), 1217.

Sidanius, J., Feshbach, S., Levin, S., & Pratto, F. (1997). The interface between ethnic and nationalattachment: Ethnic pluralism or ethnic dominance? Public Opinion Quarterly, 61, 102133.

Simon, B. (1997). Self and group in modern society: Ten theses on the individual self and the collectiveself. In R. Spears, P. J. Oakes, N. Ellemers, & S. A. Haslam (Eds.), The social psychology of stereotyping and group life (pp. 318335). Oxford: Blackwell.

Simon, B., Kulla, C., & Zobel, M. (1995). On being more than just a part of the whole: Regional identityand social distinctiveness. European Journal of Social Psychology, 25, 325340.

Smith, E. R., & Henry, S. (1996). An in-group becomes part of the self: Response time evidence.Personality and Social Psychology Bulletin, 22, 635642.

Stangor, C., & McMillan, D. (1992). Memory for expectancy-congruent and expectancy-incongruent

information: A review of the social and social developmental literatures. Psychological Bulletin,111, 4261.

Stryker, S. (1980). Symbolic interactionism: A social structural version . Menlo Park, CA:Benjamin/Cummings.

Swann, S., & Wyer, R. S. (1997). Gender st

huddy

Documents